@massarin Great package. I think there still is something that is critical for this data to be useful in the long term. I propose we include all columns in the WASH mart. Add all code columns, for example Loc which has ISO 3166-1 alpha-3 codes and all the columns with indicator codes and the underlying values for text values such as in the Dimension columns.

Why?

In real world data analysis I find myself regularly depending on the codes for countries and indicators to unambiguously combine datasets. This becomes particularly important across monitoring cycles when the labels that describe codes or the names of countries may change and then it becomes more difficult to compare and sometimes unfeasible.

Proposal:

Add a helper function for an abridged and human readable table, for example glaas_simple(). The glaas object returns the entire dataset including value and code columns.

Add another data-orientated example using select explaining how select can manage the number of columns by selecting columns by name or even by a word in the column.

library(glaas)
library(tidyverse)

glaas %>% 
    filter(
        Time > 2020, 
        LocText == "Brazil", 
        IsComparable_2013 == TRUE, 
        GrandParentText=="Monitoring", 
        Dim1ValText == "Sanitation"
    ) %>% 
    select(
        Time, 
        Loc,
        IndText_HL, 
        contains("Parent"), 
        contains("Text")
    ) %>% 
    arrange(Dim3ValText) %>% 
    View()